Abstract Introduction One of the major neuropathological features of Alzheimer's disease (AD) is the accumulation of amyloid‐β (Aβ) protein in the brain. Evidence suggests that the low‐density lipoprotein receptor‐associated protein (RAP) binds strongly to Aβ and enhances its cellular uptake and that decreased RAP expression correlates with increased Aβ production in animal models of AD. Methods The current study examined whether RAP levels change in AD human brain tissue and whether they are related to the amount of AD pathology. RAP and NeuN levels were determined by Western blot, while low‐density lipoprotein receptor‐related protein 1 (LRP1), tau and Aβ levels were determined by ELISA in the temporal cortex of 17 AD and 16 control cases. Results An increase in total Aβ and insoluble and soluble tau protein was observed in AD brain tissue. In contrast, RAP levels were significantly decreased in AD brain tissue compared to controls. Correlation analysis revealed that levels of RAP correlated with both total Aβ and soluble and insoluble tau levels. Neither LRP1 nor NeuN levels were significantly altered in AD brain tissue homogenates and did not correlate with Aβ or tau protein levels. Conclusion Reduction in RAP may contribute to the accumulation and aggregation of Aβ in the AD brain.

Soluble oligomers are now thought to be more toxic than monomeric Aß or insoluble Aß fibrils in Alzheimer's disease (AD). However, the relative amounts and comparative toxicity of oligomeric Aß40 versus Aß42 has not yet been determined in sporadic AD. Aß40 and Aß42 peptide concentrations were assessed by ELISA in soluble monomeric, soluble oligomeric and insoluble protein fractions extracted from the inferior temporal cortex of 10 sporadic AD and 10 aged-matched controls. For in vitro experiments homogenous preparations of oligomeric and fibrillar Aß42 and Aß40 were prepared according to published methods. Human primary fetal neurons were exposed to 10μM of each peptide for 16 hours. Conditioned media was assessed for release of lactate dehydrogenase as an indicator of toxicity. Total Aß40 (49±2ng/g protein) levels were significantly lower than Aß42 (262±35ng/g protein) in aged controls (p=0.004). In sporadic AD there were substantial increases in both soluble (70.5x p=0.03) and insoluble (21.9x p=0.01) Aß40 oligomers compared with controls who had limited Aß40 deposition. A similar increase over aged controls was not observed for Aß42 (monomer p=0.6; soluble oligomer p=0.2; insoluble p=0.2) with a proportion of aged controls containing Aß42 plaques. The AD-specific increase in Aß40 but not Aß42 in sporadic AD also dramatically increased the ratio of Aß40:Aß42 for all peptide conformations (p<0.05). In vitro studies demonstrated equivalent toxicity for oligomeric Aß40 (67±7% of neurons remained viable), oligomeric Aß42 (55±5% of neurons remained viable) and fibrillar Aß42 (62±11% of neurons remained viable) following treatment with 10μM Aß for 16 hours (Mann-Whitney U p<0.04). Fibrillar Aß40 was not toxic to neurons. Overall these data show that a neurotoxic build up of soluble Aß40 oligomers occurs in sporadic AD, suggesting that maintenance of low quantities of this peptide is required for protection against the disease process.

Alzheimer's disease is characterised by the accumulation of insoluble Aβ peptides in the cortex as extracellular plaques, in association with neurofibrillary tangles and significant neuronal loss. Inflammation is also a significant feature of the disease process and is associated with greater neurotoxicity. Formalin-fixed tissue sections from the inferior temporal cortex of 6 sporadic AD and 6 age-matched controls was used for analysis. Triple labelling immunohistochemistry using either Aβ40 and Aβ42 specific antibodies along with antibodies against reactive microglia (HLA-DR) and astrocytes (GFAP) was used to localise and quantify Aβ40 and Aβ42 to plaque pathology and determine their association with inflammation as a measure of cellular toxicity. The oligomeric-specific antibody, A11, was also double labelled with Aβ40, Aβ42 or HLA-DR imunohistochemistry to localise this Aβ conformation and determine its association with inflammation in situ. Analysis of plaque pathology demonstrated significant increases in sporadic AD in both Aβ40 and Aβ42 immunopositive plaque load (Aβ40 p= .01; Aβ42 p=.001) and density (Aβ40 p= .008; Aβ42 p=.001) compared to age-matched controls. The largest and most AD-specific change was the Aβ40 plaque load (33-fold increase), density (infinite-fold increase) and toxicity (84-97% of plaques).These Aβ40 positive plaques predominantly contained oligomeric species and were associated with a toxic glial infiltrate. Overall these data show that Aβ42 plaques occur in non-demented cases and have lower toxicity in AD. In contrast, soluble Aβ40 oligomers appear more specific to AD and are associated with greater inflammatory toxicity.

Frontotemporal lobar degeneration (FTLD) is a common cause of presenile dementia characterised by behavioural and language disturbances. Pick's disease (PiD) is a subtype of FTLD, which presents with intraneuronal inclusions consisting of hyperphosphorylated tau protein aggregates. Although Alzheimer's disease (AD) is also characterised by tau lesions, these are both histologically and biochemically distinct from the tau aggregates found in PiD. What determines the distinct characteristics of these tau lesions is unknown. As phosphorylated, soluble tau has been suggested to be the precursor of tau aggregates, we compared both the level and phosphorylation profile of tau in tissue extracts of AD and PiD brains to determine whether the differences in the tau lesions are reflected by differences in soluble tau. Levels of soluble tau were decreased in AD but not PiD. In addition, soluble tau was phosphorylated to a greater extent in AD than in PiD and displayed a different phosphorylation profile in the two disorders. Consistently, tau kinases were activated to different degrees in AD compared with PiD. Such differences in solubility and phosphorylation may contribute, at least in part, to the formation of distinct tau deposits, but may also have implications for the clinical differences between AD and PiD.

Familial Alzheimer's disease due to presenilin I (PSENI) mutations shows considerable phenotypic variability with differences in neuropathology and neurological symptoms. Spastic paraparesis is a common neurological phenotype associated with Alzheimer's disease arising from PSENI mutations. To investigate whether known genes that cause spastic paraparesis could act as Alzheimer's diseasemodifier genes, we sequenced nine spastic paraparesis genes in three Alzheimer's disease families with PSENI exon 9 deletions. We did not observe any correlation of polymorphisms or mutations in the nine spastic paraparesis genes with the variable phenotype seen in families with Alzheimer's disease and spastic paraparesis. These results suggest a need for a continuing search for genes that cause the phenotypic variation in Alzheimer's disease and spastic paraparesis.

Interest in the beta amyloid (Abeta) peptides continues to grow due to their known accumulation in the brains of patients with Alzheimer's disease and recent tantalising evidence that reducing such accumulations can reverse disease-associated functional deficits. Abeta peptides are naturally produced in every cell by proteolytic cleavage of the amyloid precursor protein with two main alloforms (40 or 42 amino acids) both of which are disease associated. The identification that genetic mutations causing Alzheimer's disease impact on Abeta production and clearance have allowed for the manipulation of these pathways in cellular and animal models. These studies show that the amount and type of Abeta in the brain has significant consequences on neural function. However, there have been significant difficulties in the conversion of these findings into successful treatments in humans. In this review we concentrate on data from human studies to determine any comparative differences in Abeta production and clearance that may assist with better treatment design and delivery. Abeta(40) is the dominant peptide species in human cerebrospinal fluid accounting for approximately 90% of total Abeta under normal conditions. However, similar studies using disease free human brain tissue do not correlate with these findings. In these studies, concentrations of Abeta(40) are low with Abeta(42) often identified as the dominant species. The data suggest preferential brain tissue utilisation and/or clearance of Abeta(40) compared with Abeta(42), findings which may have been predicted by their physiochemical differences. In Alzheimer's disease this equilibrium is disrupted significantly increasing Abeta peptide levels in brain tissue. The disease-specific increase in Abeta(40) brain tissue levels in Alzheimer's disease appears to be an important though overlooked pathological change compared with the well-documented Abeta(42) change observed both in the aged and in Alzheimer's disease. These findings are discussed in association with Abeta peptide function and a model of toxicity developed.

Inflammation, in the form of reactive astrocytes and microglia, is thought to play an important role in Alzheimer's disease (AD) pathogenesis where it correlates with brain atrophy and disease severity. The Aβ protein, which comprises senile plaques, is thought to be responsible for initiating this inflammatory response. Despite having a more aggressive disease process and greater Aβ deposition, few studies have investigated inflammation in early onset AD cases with mutations in the presenilin‐1 (PS‐1) gene. In fact, many researchers place importance on a variant plaque pathology in PS‐1 cases, known as cotton wool plaques, which lack significant inflammatory infiltrate. We investigated the association between inflammation and plaque pathology in PS‐1 AD. Classic cored, cotton wool and diffuse Aβ plaques were observed in all cases. PS‐1 cases also exhibited a novel plaque pathology with a significantly greater inflammatory response in the form of reactive microglia and astrocytes. These ‘inflammatory plaques’ consisted of a dense cresyl violet‐, silver‐, and thioflavin S‐positive, but Aβ‐, tau‐, apolipoprotein E (ApoE)‐, non‐Aβ component of Alzheimer's disease amyloid (NAC)‐ and PS‐1‐negative core. These findings indicate potent stimulator(s) of inflammation that are not typical of the Aβ that accumulates in the pathological hallmarks of sporadic AD. Identification of this substance may be important for the development of future therapeutic strategies.

Multiple degenerative hallmarks characterize Alzheimer's disease: insoluble protein deposition, neuronal loss and cortical atrophy. Atrophy begins in the medial temporal lobe and becomes global by end stage. In a small proportion of cases, these tissue changes are caused by mutations in three known genes. These cases are affected earlier in life and have more abundant protein deposition, which may indicate greater tissue atrophy and degeneration. This issue remains unresolved. Grey matter atrophy in different cortical regions was determined in genetic cases of Alzheimer's disease (N = 13) and compared to sporadic cases (N = 13) and non-diseased controls (N = 23). Genetic mutations were found to influence the degree and regional pattern of atrophy. The majority of cases had greater medial temporal atrophy than sporadic disease, suggesting that abnormalities affecting Abeta metabolism selectively increase hippocampal degeneration. Cases with mutations in presenilin-1 demonstrated additional increased frontotemporal atrophy. This effect may be due to the influence of presenilin-1 on tau phosphorylation and metabolism. These differences may explain the earlier onset ages in these different forms of Alzheimer's disease.

Mutations in presenilin-1 (PS-1) account for the majority of familial Alzheimer's disease (AD). While increasing Abeta42 is one mechanism whereby PS-1 mutations are thought to exert their pathogenic effect, little is known about the role of tau in PS-1 AD. This study compares staining (AT8 and tau-2), morphology and quantity of tau-immunoreactive cortical plaques in six PS-1 and five sporadic AD cases. The densities of tau-positive plaques differentiated PS-1 from sporadic AD cases. All PS-1 cases demonstrated a greater than 6-fold increase in tau-2-positive plaques. In PS-1 cases with mutations in exons 5 and 6, there was an increase in classical AD plaques containing hyperphosphorylated tau (AT8- and tau 2-positive). However, cases with exon 8 and 9 mutations had numerous cotton wool plaques containing nonhyperphosphorylated tau (tau-2-positive, AT8-negative). These findings suggest that PS-1 mutations increase tau deposition while mutation-specific cellular responses determine phosphorylation events and may influence cell death mechanisms.